10 Replies to “Obviating the Mood, but Very Much under Control: An Analysis of Spanish Heritage Speakers’ Knowledge of Binding Constraints (Silvia Perez-Cortes)”

  1. Amazing presentation! Thank you for sharing this.

    At the end of the talk, you give us a sneak preview of the production data. I wonder whether you’ve thought about using participants’ production patterns as a predictor of their interpretation patterns, e.g., by using their proportion of querer que + subjunctive production as a continuous predictor.

    If you show that heritage speakers who use querer + que + subjunctive less are more innovative in their comprehension patterns, I think that would be super compelling evidence of the link that you seem to be pointing to in your framing of the study.

    Thanks again!

    1. Thanks for your comment, David!

      That is definitely my thought process! I wish I’ve had more time to run the analysis so that I could connect the two (hinting at the potential effects of optionality was all I could do), but this is certainly where I am heading. Something else that I want to explore, is whether HSs tended to produce the embedded subject explicitly in the production task -and compare it to controls-, and if doing so (or not) had an effect on 1) the likelihood of producing subjunctive in the subordinate clause of disjoint reference desideratives (querer +que+ ?); 2) the interpretation of the null embedded subject in subjunctive complements.

      Comparing individual data across two two tasks could also be a game-changer, as it would point to the types of tendencies that you mentioned at the end of your comment 🙂

      Thanks so much for watching!

      1. Thanks for a fascinating presentation, Silvia. What interesting results you found! I will also be very interested to hear what you find looking at the relationship between this data and the production data. Based on the sneak preview you presented, it sure looks like you found very similar trends with that data. I think David’s suggestion for an analysis is a very good one (and sounds like an approach you were already considering). I’m not sure if the graph you showed quickly for production includes all the same participants (as I know that data analysis is in progress), but based on that graph, I would think that there is a good chance you would find what David mentioned, which would be strong evidence for the link, and very cool! Also, I love the idea you mentioned in your follow-up comment about looking at HSs production of explicit subjects in the embedded clause. Based on research that is out there (also with L2 learners of Spanish coming from L1 English), I would bet you’d find higher rates of explicit subject use in the subordinate clause as compared to controls, and it will be very interesting to see if the presence of that explicit subject increases the likelihood of using the subjunctive, as you mentioned. I would predict that it very well could. I don’t have as strong as a prediction regarding your second idea looking at a possible connection with the interpretation of the null subject, but I can’t wait to hear what you discover. Keep me posted please! Thanks again for a wonderful presentation, as always.

  2. What a great presentation!

    I just wanted to say that I look forward to your study with production data and also the one with other predicates. The examples you gave throughout the presentation made me wonder about how different the results could be if we compare “querer que” and “no creer que”, for example. It would be interesting to see if a linguistic factor (maybe lexical identity of the governor, clause type or negation) would be significant as well.
    Regarding the sneak preview of the production data at the end, I was wondering if you could share with us which alternative constructions you have found and categorized as “Other”.

    Thank you so much! ?

    1. Thanks for your comment, Isabella!

      Regarding your first point, I think one of the biggest issues when other predicates are examined (when we look at mood selection or any other epiphenomena related to the presence of subjunctive) is that we need to take into account the type of modality conveyed by the modal form as well as the moment when said constructions are acquired/developed. In the case of disjoint-reference desideratives like querer, for example, acquisition of this type of deontic structures occurs rather early (between 3-5, although there is a period of optionality as well), whereas non-assertive constructions are not mastered until much later (6/7-9) even in monolingual populations. This is something that we need to take into account when we analyze the performance of HSs, as their input and activation of the HL is crucially reduced when some of these structures are still being developed.

      Regarding your other comment (and thanks for asking, I *love* talking about production data!) there were MANY different constructions used by the participants (most of which appear in my paper Perez-Cortes, Putnam & Sánchez (2019) Differential Access: Asymmetries in Accessing Features and Building Representations in Heritage Language Grammars ). Other than the typical querer + que + indicative, there were also instances of querer que + infinitive (“Quiere que ir…..”), querer que + periphrasis of obligation (“Quiere que puede ir…”), or a sort of English-like ECM construction (“Quiere Maria ir….”) amongst others 🙂 The interesting thing is that examples of divergences increased in % and also in variety as HSs’ proficiency in Spanish decreased, reinforcing the idea that proficiency modulates morphological variability among HSs.

      Thanks again for your questions!

      1. Thank you for your reply! I’ll definitely check out this paper and your future work! ?

  3. Dear Silvia,
    Excellent study and very interesting results. I find the effect of proficiency very interesting, especially in line of what these speakers are “learning” about the use of SUBJ with increasing proficiency (or the implications of the use of SUBJ), namely non-correferentiality. Maybe with more proficiency they are “getting” first the type of relationship between referents that is most common in the language in association with the SUBJ, i.e., non-co-referentiality. I wonder if this then requires some re-structuring or extra processing efffort when trying to use / comprehend co-referential SUBJ cases. I also wonder if we were to examine processing of those less frequent SUBJ co-referential sentences in monolingual speakers whether we would find that they also entertain strong competition from the most frequent non-co-referential use associatd with SUBJ. Just thougts that I had while listening to this great presentation. Thanks!

    1. ¡Buenos días Pablo! Thanks for your comment.

      One of the things I wanted to explore with this project was the fact that while mood (SUBJ) selection within this type of deontic predicates has usually been labelled as one the most stable ones among HS due to the categorical use of subjunctive, there is an extra layer of complexity associated to the presence of this form that might/might not have been acquired. I agree with you that HSs with higher levels of proficiency in the HL most likely have an advantage when integrating the syntactic/semantic features associated with non-corref desiderative constructions onto the expected morphological exponents.

      The point about co-referential SUBJ potentially being more costly to process (or even the fact that some sort of “reanalysis” would have to occur in order to accommodate for these cases) is super interesting, and something that I had not thought of before, thanks!!
      I think this goes along with my comment to Isabella above, in the sense that, it would most likely increase the chances of observing optionality (both at the level of interpretation and comprehension) in cases like non-assertive predicates (No creo que…) where not only there is the possibility of an alternation between IND/SUBJ, but also the presence of many binding relations to be considered wrt to the matrix/subordinate clauses.

      I do think that in monolingual populations we’d most likely find a strong competition between the non-coreferential uses of SUBJ and co-ref ones in these last set of predicates. What’s interesting for me would be to see whether the same set of alternatives would be entertained by both groups (control and HSs) or if there’d be a restriction of the possibilities to be considered by the latter 🙂

      Thanks so much again for your insight!

  4. Hello Silvia,

    Thanks for such a wonderful presentation. The findings of the study are very interesting and I look forward to reading about your research with SUBJ. I was just wondering if you are planning to incorporate other factors (apart from proficiency) in the analysis of your upcoming project (with production data).

  5. Hola Silvia,
    Great presentation!
    I would love to learn a bit more about the participants from both groups.
    In the case of your control group, where are they from? Are they also bilingual in Spanish and English?
    For the heritage speakers, where are their families from?
    Thanks!

Comments are closed.